The aim of this thesis was to investigate the development of pigs that were raised in a newly developed group housing system for lactating sows and their litters, with a focus on the transition around weaning and performance later in life. The starting point of our multi-suckling (MS) system was the natural behaviour of pigs, and the system consisted of 5 farrowing pens connected to a communal area, which included a communal feeding area. Litters were grouped at 1 week of age. Compared with a conventional farrowing system (in which a sow and her piglets are housed in a pen in which the sow is confined individually in a crate), the MS system provided a more spacious and complex environment, with more social and physical enrichment. Before weaning, we found that MS piglets showed more feed-directed behaviour and less damaging behaviours, such as tail biting, than conventionally housed piglets. After weaning at 4 weeks of age, when housed in a more physically and socially enriched pen, MS-raised piglets showed a higher feed intake, a higher weight gain, more play behaviour, and less maladaptive behaviour than the conventionally raised piglets that were housed in a standard nursery pen. In a follow-up study, in which all pigs were housed under equal and enriched conditions after weaning, we found that piglets raised in the MS system had a higher feed utilisation and a lower carbohydrate absorption in an oral sugar absorption test in the early post-weaning phase. In a subsequent study, the same animals were used to investigate their social and cognitive development, using behaviour tests during which pigs competed for access to feed. We found few differences, but there were indications that the pigs raised pre-weaning in the MS system made more use of social information than the conventionally raised pigs. Lastly, we studied effects of different weaning procedures in the MS system (gradual weaning during 9 weeks of lactation vs. abrupt weaning at 4 weeks of age). Weaning seemed to have less impact for the pigs subjected to the gradual weaning treatment (reflected in weight gain and maladaptive behaviour). Also on the long term during the finishing phase (during which both groups were housed equally from 9 weeks of age), the pigs subjected to the gradual weaning treatment showed less maladaptive behaviour and had fewer body lesions than abruptly weaned pigs. To conclude, the multi-suckling system seems promising for improving pig performance, behaviour, and welfare, especially in combination with a more gradual weaning procedure during an extended lactation.